
import pandas as pd
import numpy as np

import seaborn as sn
import plotly.express as px
import plotly.offline as pyo
import matplotlib.pyplot as plt


imdb = pd.read_csv('imdb.csv')
imdb.head(5)


imdb.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 20789 entries, 0 to 20788
Data columns (total 14 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   id              20789 non-null  object 
 1   name            20789 non-null  object 
 2   year            20789 non-null  object 
 3   rating          20789 non-null  float64
 4   certificate     20789 non-null  object 
 5   duration        20789 non-null  object 
 6   genre           20789 non-null  object 
 7   votes           20789 non-null  object 
 8   gross_income    20789 non-null  object 
 9   directors_id    20789 non-null  object 
 10  directors_name  20789 non-null  object 
 11  stars_id        20789 non-null  object 
 12  stars_name      20789 non-null  object 
 13  description     20789 non-null  object 
dtypes: float64(1), object(13)
memory usage: 2.2+ MB


imdb.isnull().sum()

id                0
name              0
year              0
rating            0
certificate       0
duration          0
genre             0
votes             0
gross_income      0
directors_id      0
directors_name    0
stars_id          0
stars_name        0
description       0
dtype: int64


imdb['genre'].nunique()

919


Genre = set()                     # Set to store unique elements only

# Store unique genre to Genre set()
for gre in imdb['genre']:
    l = gre.replace(',','').split()
    for item in l:
        Genre.add(item)
len(Genre)

26


vote = []

for value in imdb['votes']:
    vote.append(int(value.replace(',','')))      # Replace comma with empty space and change data type to integer

imdb['votes'] = vote
imdb['votes'].dtype

dtype('int64')


imdb.sort_values(['votes'],ascending=False)[['name','votes']][:3]


Max = imdb['votes'].max()
imdb[imdb['votes'] == Max][['name']]


imdb.groupby('genre', as_index=False).count().sort_values(by = 'name',ascending=False)[['genre','name']].head(3)


imdb.groupby('genre').size().sort_values(ascending=False).reset_index(name='Movies_counts').head(3)


# Create a dictionary with unique Genre key and 0  as values
GENRE ={}
for gr in Genre:
    GENRE[gr] = 0

# Add the count to values of respective GENRE key
for gre in imdb['genre']:
    l = gre.replace(',',' ').split()
    for gr in l:
        GENRE[gr] +=1

# Create Dataframe from GENRE dictionary key and values
Genres = pd.DataFrame(list(zip(GENRE.keys(), GENRE.values())),columns =['Genre', 'Movies_Counts'])


Genres.sort_values('Movies_Counts', ascending=False).head()


# Adjust the size and view of plot
plt.figure(figsize=(12,5), dpi=523)                                   

# Quick bar plot with pandas
imdb.genre.value_counts()[:25].plot(kind='bar', color = 'g')          

# Set horizontal grid lines
plt.grid(True, axis = 'y',                                            
         color = 'gray', 
         linestyle = '--',  
         linewidth=0.2) 
# Remove vertical grid lines
plt.grid(False, axis = 'x')      

# Remove boundry line, set Offset & trim the corner
sn.despine(offset=10, trim= True)  

# Rotate the x tick value by 90 degree
plt.xticks(rotation = 90)                                             

plt.xlabel("Top 25 Genre's Pairs", fontsize = 13)                           # Set X label         
plt.ylabel("Movies Count", fontsize = 11)                                   # Set Y label
plt.title("Top 25 Genre's Pairs w.r.t Movies counts", fontsize = 15)        # Set title
plt.show()


fig =px.bar(imdb[['genre']].value_counts().reset_index(name='Movies Counts')[:25],
            x= 'genre',
            y= 'Movies Counts',
            labels= {'genre':"Genre's Pairs"},
            template='plotly_dark',
            title="<b> Top 25 Genre's Pairs w.r.t Movies counts")

fig.show()

pyo.plot(fig, filename = "Plotly/Top 25 Genre's Pairs w.r.t Movies counts.html", auto_open = True)

"Plotly/Top 25 Genre's Pairs w.r.t Movies counts.html"


# Adjust the size and view of plot
plt.figure(figsize=(15,5), dpi=523)                                       
#sn.set_style('whitegrid') 

# Seaborn Bar plot
sn.barplot(data = Genres.sort_values('Movies_Counts', ascending=False),   
           x='Genre',
           y = 'Movies_Counts',
           palette= "crest_r"#'plasma'
          )     


# Set horizontal grid lines
plt.grid(True, axis = 'y',                                                 
         color = 'gray', 
         linestyle = '--',  
         linewidth=0.2
        ) 
# Remove vertical grid lines
plt.grid(False, axis = 'x')                                   

# Remove boundry line, set Offset & trim the corner
sn.despine(offset=10, trim=False)

# Rotate the x tick value by 90 degree for better view
plt.xticks(rotation = 90)                                     

plt.xlabel("Genre", fontsize = 13)                             # Set X label         
plt.ylabel("Movies Count", fontsize = 11)                      # Set Y label
plt.title('Movies counts as per Genre', fontsize = 16)         # Set title
plt.show()


fig =px.bar(Genres.sort_values('Movies_Counts', ascending=False),
            x= 'Genre',
            y= 'Movies_Counts',
            template='plotly_dark',
            color_discrete_sequence = px.colors.qualitative.Dark2,
            title='<b>Movies counts as per Genre')

fig.show()

pyo.plot(fig, filename = "Plotly/Movies counts as per Genre.html", auto_open = True)

'Plotly/Movies counts as per Genre.html'


gross = []
for income in imdb['gross_income']:
    income = income.replace(',','')                  # Replace comma with whitespace
    try:
        gross.append(int(income))                    # Convert to int and add to the list
    except:
        income = income.replace('$','')              # Replace $ symbol with empty space
        income = income.replace('M','')              # Replace M (i.e Million) with empty space
        inc = float(income)*(10**6)                  # Convert to float and Multiply by 1Million(10**6)
        gross.append(int(inc))                       # Convert to integer and append in the existing list

imdb['gross_income'] = gross       # Repllace the existing gross_income column with newly gross list in existing dataframe
imdb['gross_income'].dtypes        # Check the data type of gross_income column

dtype('int64')


df = imdb.sort_values('gross_income')[['name','gross_income']].sort_values(by = 'gross_income',ascending=False)[:10]


# Adjust the size and view of plot
plt.figure(figsize=(15,5), dpi=523)                                       
#sn.set_style('whitegrid') 

# Set Y limit value 
plt.ylim(600000000, 950000000)                               

# Seaborn Bar plot
sn.barplot(x=df.name.values,           
           y=df.gross_income.values,
           data=df,
           #color = 'green',
           palette= "crest_r"
          )  


# Set horizontal grid lines
plt.grid(True, axis = 'y',                                                 
         color = 'gray', 
         linestyle = '--',  
         linewidth=0.2
        ) 
# Remove vertical grid lines
plt.grid(False, axis = 'x')                                   

# Remove boundry line, set Offset & trim the corner
sn.despine(offset=10, trim=False)

# Rotate the x tick value by 90 degree for better view
plt.xticks(rotation = 90)                                     

plt.xlabel("Movies Name", fontsize = 13)                             # Set X label         
plt.ylabel("Gross income (10^8)", fontsize = 11)                     # Set Y label
plt.title('Top 10 Movies by Gross income', fontsize = 15)            # Set title
plt.show()


df = imdb.sort_values('gross_income')[['name','gross_income']].sort_values(by = 'gross_income',ascending=False)[:10]

fig =px.bar(df,
            x       = 'name',
            y       = 'gross_income',
            color   = 'name',
            labels  = {'name': 'Movies Name'},
            color_discrete_sequence = px.colors.qualitative.Dark2,
            template='plotly_dark',
            title   ='<b> Top 10 Movies by Gross-income'
           )

fig.show()
pyo.plot(fig, filename = "Plotly/Top 10 Movies by Gross-income.html", auto_open = True)

'Plotly/Top 10 Movies by Gross-income.html'


time = []
for t in imdb['duration']:
    T = t.replace(',','').split()           # Replace comma with whitespaces and split     
    try :
        time.append(int(T[0]))              # Add the integer value to the list
    except:
        print(t)

imdb['duration'] = time                     # Replace the existing duration with newly created list
imdb['duration'].dtypes                     # Check the data type

dtype('int64')


df = imdb.groupby('genre', as_index=False)[['duration']].mean().sort_values(by = 'duration',ascending=True)
df.head(5)


Duration ={}

# Create a Dictionary with empty list
for gr in Genre:                                         
    Duration[gr] = []

# Add movies duration time for each Genre respectively     
for i in range(imdb.shape[0]):                             
    l = imdb.genre.iloc[i].replace(',',' ').split()           # Replace comma and split
    for gr in l:
        Duration[gr].append(imdb.duration[i])                # Add all the movie duration for the respective Genre


# Replace the existing Duration list with the mean Duration time
for key in Duration.keys():                             
    Duration[key] = sum(Duration[key])/len(Duration[key])

# Canvert the Dictionary to DataFrame
DURATION = pd.DataFrame(list(zip(Duration.keys(), Duration.values())),columns =['Genre', 'Durations'])

# Add DURATION to the existing Genres DataFrame as Avg_Duration
Genres['Avg_Duration'] = DURATION['Durations']


Genres.sort_values('Avg_Duration', ascending=True)[['Genre','Avg_Duration']]


fig =px.bar(Genres.sort_values('Avg_Duration', ascending=True)[['Genre','Avg_Duration']],
            x= 'Genre',
            y= 'Avg_Duration', 
            labels={'Genre':'Genre'},
            color = 'Genre',
            color_discrete_sequence = px.colors.sequential.Viridis,
            template='plotly_dark',
            title='<b> Average Movies duration (in Minutes) w.r.t. Genre')

fig.show()
pyo.plot(fig, filename = "Plotly/Average Movies duration (in Minutes) w.r.t. Genre.html", auto_open = True)

'Plotly/Average Movies duration (in Minutes) w.r.t. Genre.html'


df = imdb.groupby(['genre'], as_index=False)[['gross_income']].mean().sort_values(by = 'gross_income',ascending=False)
df.head()


Income ={}

# Create a Dictionary with empty list
for gr in Genre:                                         
    Income[gr] = []

# Add movies Gross Income for each Genre respectively in a list    
for i in range(imdb.shape[0]):                             
    l = imdb.genre.iloc[i].replace(',',' ').split()    # Replace the comma and split
    for gr in l:
        Income[gr].append(imdb.gross_income[i])        # Add all the gross_income for their respectly unique Genre


# Replace with existing Gross Income list with the mean of Gross Income
for key in Income.keys():                             
    Income[key] = sum(Income[key])/len(Income[key])

# Canvert the Dictionary to DataFrame
Incomes = pd.DataFrame(list(zip(Income.keys(), Income.values())),columns =['Genre', 'Avg_gross_income'])

# Add Gross Income to the existing Genres DataFrame
Genres['Avg_gross_income'] = Incomes['Avg_gross_income']


Genres.sort_values(['Avg_gross_income'],ascending=False).head(3)


# Adjust the figure size and view
plt.figure(figsize=(13,5),dpi=523)                           

# Seaborn Bar plot
sn.barplot(x='Genre',                                        
           y='Avg_gross_income',
           data=Genres,
           order = Genres.sort_values(['Avg_gross_income'],ascending=False)['Genre'].values,
           #color = 'green',
           palette= "crest_r"
          )

# Remove boundry line,set offset at corner & trim the corner
sn.despine(offset=10, trim= True)                              

# Rotate the x tick value by 90 degree for better view
plt.xticks(rotation = 90)              

plt.xlabel("Genre", fontsize = 13)                                    # Set X label         
plt.ylabel("Gross income (10^7)", fontsize = 11)                      # Set Y label
plt.title('Mean Gross Income w.r.t to Genre', fontsize = 15)          # Set title
plt.show()


Genres[Genres['Avg_gross_income']==Genres['Avg_gross_income'].max()]


df = imdb.groupby(['genre'], as_index=False)[['gross_income']].mean().sort_values(by = 'gross_income',ascending=False)
df.tail()


fig =px.bar(Genres.sort_values(['Avg_gross_income'],ascending = True),
            x= 'Genre',
            y= 'Avg_gross_income', 
            labels={'Genre':'Genre'},
            color  = 'Genre',
            color_discrete_sequence = px.colors.sequential.Cividis_r,
            template='plotly_dark',
            title='<b> Average Gross Income w.r.t. to Genre')

fig.show()
pyo.plot(fig, filename = "Plotly/Average Gross Income w.r.t. to Genre.html", auto_open = True)

'Plotly/Average Gross Income w.r.t. to Genre.html'


Genres[Genres['Avg_gross_income']==Genres['Avg_gross_income'].min()]


# Adjust the figure size and view
plt.figure(figsize=(12,5), dpi=223)

#  Average Gross Income distribution plot
sn.histplot(data = imdb.groupby('genre')[['gross_income']].mean(),
            #bins = 26,
            stat = 'count',                                 
            kde = True,
            log_scale=10,
           )

# Remove boundry line,set offset at corner & trim the corner
sn.despine(offset=10, trim= False)                            

plt.xlabel("Average Gross Income", fontsize = 11)                                  # Set X label         
plt.title("Average Gross Income Ditribution w.r.t Genre's Pairs", fontsize = 13)   # Set title
plt.show()


fig =px.scatter(imdb,
                x= 'name',
                y= 'gross_income',
                labels={'name':'Movies'},
                size='gross_income',
                size_max= 35,
                color = 'genre',
                color_discrete_sequence = px.colors.sequential.Plasma,
                opacity=0.8,
                template='plotly_dark',
                title="<b> Average Gross Income Ditribution w.r.t Genre's Pairs")

fig.update_layout(autosize=True,
                  height=800,
                  xaxis=dict(showgrid=False),
                  yaxis=dict(showgrid=False)
                 )

fig.show()
pyo.plot(fig, filename = "Plotly/Average Gross Income Ditribution w.r.t Genre's Pairs.html", auto_open = True)

"Plotly/Average Gross Income Ditribution w.r.t Genre's Pairs.html"


plt.figure(figsize=(12,5), dpi=223)
sn.histplot(data = Genres,
            x = 'Avg_gross_income',
            #bins = 26,
            stat = 'count',
            kde = True,
            log_scale=10,
           )

# Remove boundry line,set offset at corner & trim the corner
sn.despine(offset=10, trim= False)                            

plt.xlabel("Average Gross Income", fontsize = 12)                            # Set X label         
plt.title("Average Gross Income Ditribution w.r.t Genre", fontsize = 12)     # Set title
plt.show()


fig =px.histogram(Genres,
                x= 'Genre',
                y= 'Avg_gross_income',
                labels={'name':'Genre'},
                #size='Avg_gross_income',
                color = 'Genre',
                color_discrete_sequence = px.colors.sequential.Darkmint,
                template='plotly_dark',
                title='<b> Average Gross Income Distribution w.r.t Genre')

fig.update_layout(autosize=True,
                  xaxis=dict(showgrid=False),
                  yaxis=dict(showgrid=False)
                 )

fig.show()
pyo.plot(fig, filename = "Plotly/Average Gross Income Distribution w.r.t Genre.html", auto_open = True)

'Plotly/Average Gross Income Distribution w.r.t Genre.html'


df = imdb.groupby('genre', as_index=False).count().rename(columns = {'name':"Movie's Count"})[['genre',"Movie's Count"]]
df.head()


Genre_Pairs = {}
# Add the movie count of genre 2 pairs to Genre_Pairs
for i  in range(df.shape[0]):
    l = df['genre'][i].replace(',',' ').split()
    if len(l)==2:
        Genre_Pairs[df['genre'][i]] = df["Movie's Count"][i]
        #print(df['genre'][i],'-->',df["Movie's Count"][i])

# Create Dataframe from Genre_Pairs dictionary key and values
Genre_2Pairs = pd.DataFrame(list(zip(Genre_Pairs.keys(), Genre_Pairs.values())),columns =['Genre 2 Pairs','Movies_Counts'])


Genre_2Pairs.sort_values('Movies_Counts', ascending=False).head()


df = imdb.groupby('genre', as_index=False).count().rename(columns = {'name':"Movie's Count"})[['genre',"Movie's Count"]]
df.sort_values("Movie's Count", ascending=False).head(2)


Genres.sort_values('Movies_Counts', ascending=False)[['Genre','Movies_Counts']].head(2)

	id	name	year	rating	certificate	duration	genre	votes	gross_income	directors_id	directors_name	stars_id	stars_name	description
0	tt0118708	Beverly Hills Ninja	(1997)	5.5	PG-13	88 min	Action, Comedy	41511	31,235,710	nm0240797	Dennis Dugan	nm0000394,nm0001733,nm0795225,nm0662511	Chris Farley,Nicollette Sheridan,Robin Shou,Na...	A man tries to rescue a woman with a little he...
1	tt1789846	Monster High: New Ghoul at School	(2010 TV Movie)	6.7	TV-Y7	23 min	Animation, Fantasy	567	11,000	nm0655637,nm0705779	Audu Paden,Eric Radomski	nm1512156,nm1086483,nm0220635,nm1154161	Kate Higgins,Salli Saffioti,Debi Derryberry,La...	Frankie Stein begins at Monster High and follo...
2	tt12204610	The Killer in the Guest House	(2020 TV Movie)	5.3	TV-14	85 min	Thriller	234	132,863	nm1046827	Tony Dean Smith	nm0387581,nm4785386,nm2254636,nm5512848	Chelsea Hobbs,Marcus Rosner,Matthew Kevin Ande...	A struggling fashion photographer unknowingly ...
3	tt6708668	Black Crab	(2022)	5.6	TV-MA	114 min	Action, Adventure, Drama	18,616	31,153,464	nm3515833	Adam Berg	nm0636426,nm1939580,nm1502469,nm4558335	Noomi Rapace,Jakob Oftebro,Dar Salim,Erik Enge	In a post-apocalyptic world, six soldiers on a...
4	tt0453533	Unrest	(I) (2006)	5.0	R	88 min	Horror, Thriller	8,066	215,300	nm1118463	Jason Todd Ipson	nm0257598,nm1404049,nm1130845,nm0662475	Corri English,Scot Davis,Joshua Alba,Marisa Pe...	A young pathology med student suspects that th...

	name	votes
11722	The Dark Knight	2574832
7374	Inception	2284273
20215	Pulp Fiction	1995361

	genre	name
43	Action, Crime, Drama	741
456	Crime, Drama, Thriller	527
441	Crime, Drama	466

	genre	Movies_counts
0	Action, Crime, Drama	741
1	Crime, Drama, Thriller	527
2	Crime, Drama	466

	Genre	Movies_Counts
12	Drama	9928
14	Action	5508
0	Crime	5444
15	Comedy	4901
24	Thriller	4299

Loading the libraries and dataset¶

Data Cleaning¶

1. How many different Genre of Movies are there in this dataset?¶

2. Find the most voted Movie in the dataset¶

Approach :1¶

Approach :2¶

3. Which Genre is having most number of Movies in the dataset?¶

Approach : 1¶

Approach :2¶

For unique Genre's Pairs¶

For unique Genre¶

4. Find the Disctribution of number of movies in each genre with bar graph¶

5. Plot top 10 Movies with maximum Gross-Income with Bar Graph¶

Convert "gross_income" data type from object to Integer¶

6. Whats the average duration of a movie in each genre in ascending order?¶

With respect to Unique Genre's Pairs¶

With respect to Unique Genre¶

7. Which Genre is the best to make a movie to get maximum gross income?¶

With respect to unique Genre's Pairs :¶

With respect to unique Genre :¶

8. Which Genre is the worst to make money from a movie?¶

With respect to unique Genre's Pairs :¶

With respect to unique Genre :¶

9. Distribution of Average Gross Income of each genre?¶

For unique Genre¶

10. Which two pairs of Genre is having maximum number of movies?¶

Top two Genre pairs Group w.r.t to Movies count¶

Top two Genre w.r.t to Movies count¶

	genre	duration
600	Documentary, Short, Action	1.0
869	Short, Comedy, Fantasy	3.0
298	Animation, Short, Music	3.0
864	Short, Biography, Comedy	3.0
857	Short, Action, Fantasy	3.0

	Genre	Avg_Duration
1	Short	16.871595
9	Animation	78.081154
3	News	78.083333
8	Game-Show	85.000000
25	Film-Noir	88.260586
13	Horror	93.198355
11	Family	93.464522
4	Western	96.052500
15	Comedy	96.276678
22	Reality-TV	98.666667
7	Sci-Fi	98.919723
19	Fantasy	99.555921
16	Adventure	100.718882
24	Thriller	100.723889
5	Mystery	100.821799
18	Documentary	100.858595
23	Music	102.326308
0	Crime	102.869214
6	Sport	103.804613
2	Romance	105.266848
14	Action	105.594590
12	Drama	106.126511
20	War	107.430716
10	Biography	109.628390
21	Musical	110.493590
17	History	116.873188

	genre	gross_income
809	Music, Romance, Sci-Fi	2.097260e+08
14	Action, Adventure, Sci-Fi	1.164601e+08
76	Action, Family, Sport	1.151040e+08
236	Animation, Adventure, Comedy	8.432863e+07
270	Animation, Drama, War	8.405276e+07

	Genre	Movies_Counts	Avg_Duration	Avg_gross_income
16	Adventure	3543	100.718882	3.818729e+07
9	Animation	1109	78.081154	3.551142e+07
7	Sci-Fi	1445	98.919723	3.128869e+07

	genre	gross_income
257	Animation, Comedy, Sci-Fi	1469.0
796	Horror, Thriller, Fantasy	1027.0
860	Short, Adventure	866.0
85	Action, History, Horror	438.0
732	Family, Mystery, Fantasy	181.0

	genre	Movie's Count
0	Action	118
1	Action, Adventure	49
2	Action, Adventure, Biography	15
3	Action, Adventure, Comedy	362
4	Action, Adventure, Crime	121

	Genre 2 Pairs	Movies_Counts
58	Crime, Drama	466
94	Drama, Thriller	425
122	Horror, Thriller	328
39	Biography, Drama	272
95	Drama, War	269